Single 5" flywheel shooter and ball squishiness

My daughter and I have been doing testing with her 5" single flywheel shooter. We are getting about 75 percent of the ball in. Checking the encoder log the ball that did not went in have the same flywheel speed as the other ball. So this is not a controller issue. We quickly realize the ball that did not go in are always the same bunch of balls and they are either too hard or too soft compare with the rest.
I read some people sort the balls out to two pile before the match and use different flywheel speed but I can see that to be extremely problematic in competition.

What are people’s solution to this from a build point of view?
More compression? Less compression?
More contact time or less contact time with the flywheel?
Maybe use a small wheel instead of a compression plate/hood?
Any other tricks?

Any help appreciated.

A common issue could be that you are compressing the balls too much. The result of this is a major slow down in the speed of the flywheel, and if you load the second ball too quickly, you will then find that said shot will be much weaker. You can either load the balls slower to allow for the shooter to regain it’s maximum speed, or you can compress the balls less which will then reduce the reload time.
It seems that there are two major categories of ball density, squishy and hard. You should play around with different shot areas, and should find one that can have one set of balls hitting the top of the net, and having the other set hitting the lower part of the net (both in the high goal, as I assume this is your goal).

less contact time would be best for a single wheel. problems seem to pop up a lot when the ball has to travel along the flywheel for any distance.
look at a picture of this baseball pitcher.
[ATTACH]9724[/ATTACH]

It only has a plate for the wheel to push the ball against. You also want a good amount of pressure (more than most would guess) with any kind of flywheel shooter. All this should help accuracy, but you will never get 100%. 2 wheel flywheels are normally known for being more accurate for a lot of nitty gritty reason.

good luck!
f692e5ee707bd2a391d3df7ccc017e0f.jpg

You might be either compressing the ball to much or have an early set of balls. You can buy a set of 25 balls on the vex website.

http://www.vexrobotics.com/vexedr/products/competition-products/276-4257.html

Some other useful info can be found in some other threads discussing ball squishiness, it seems to be a bit of problem :stuck_out_tongue:

https://vexforum.com/t/inconsistencies-in-ball-density/29501/1
https://vexforum.com/t/ball-density-distance-calculations/29718/1
https://vexforum.com/t/ball-squishy-ness-increasing/30505/1
https://vexforum.com/t/squishyness-of-the-ball-is-affecting-my-shot/30549/1

Point well taken. We are well aware of the flywheel recovery time. In the beginning we just time it by counting out loud. Our encoder log told us that the recovery time is about one sec so we chanted “one missippi feed” between shots. We now have a line of code that stop the ball conveyor from feeding until the flywheel has recovered.
We are now playing around with the compression which turns out to be very sensitive. Even one washer makes a huge difference. My daughter has invented a squishiness score: RAW RIPE & ROTTEN. We are now at the point where we can get both the RIPE and ROTTEN one in but the RAW one we still can’t get them in without boosting motor power by 10.
I guess the next thing to tune is the trajectory. Does anybody have data on flat vs lobbing trajectory. Right now the exit angle for our shooter is about 50 degree.